Z ENTRALER K REDITAUSSCHUSS MITGLIEDER: BUNDESVERBAND DER DEUTSCHEN VOLKSBANKEN UND RAIFFEISENBANKEN E.V. BERLIN BUNDESVERBAND DEUTSCHER BANKEN E.V. BERLIN BUNDESVERBAND ÖFFENTLICHER BANKEN DEUTSCHLANDS E.V. BERLIN DEUTSCHER SPARKASSEN- UND GIROVERBAND E.V. BERLIN-BONN VERBAND DEUTSCHER PFANDBRIEFBANKEN E.V. BERLIN Committee of European Banking Supervisors Tower 42 (level 18) 25 Old Broad Street London EC2N 1HQ United Kingdom CP36@c-ebs.org 10th June 2010 Consultation paper CEBS Guidelines on Liquidity Cost Benefit Allocation (CP36) Dear Sir or Madam, on 10th March 2010, CEBS published the consultation paper on Guidelines on Liquidity Cost Benefit Allocation. We 1 hereby have pleasure in taking the opportunity to comment. General observations In our opinion, the Guidelines are largely in line with the general requirements for liquidity management. On the other hand, the requirement of a costs-based transfer pricing mechanism based on a risk tolerance defined specifically for each institution is more of a best practice in the liquidity risk management of the institutions. Moreover, individual requirements remain unclear or extend beyond that which is objectively required in terms of content. 1 The Zentraler Kreditausschuss (ZKA) is the joint committee operated by the central associations of the German banking industry. These associations are the Bundesverband der Deutschen Volksbanken und Raiffeisenbanken (BVR) for the cooperative banks, the Bundesverband deutscher Banken (BdB) for the private commercial banks, the Bundesverband Öffentlicher Banken Deutschlands (VÖB) for the public-sector banks, the Deutscher Sparkassen- und Giroverband (DSGV) for the savings banks financial group, and the Verband deutscher Pfandbriefbanken (VdP) for mortgage banks. Collectively, they represent more than 2,300 banks.
- 2 - With the present Guidelines, CEBS puts into concrete terms recommendation 2 of its Technical Advice on Liquidity Risk Management (CEBS 2008 147). In our estimation, this specification should not lead to tightening up the CEBS recommendation existing hitherto. For instance, in the recommendation there is a question of a transfer pricing mechanism to support the institutions. In the Guidelines, on the other hand, a transfer pricing system (e.g. in point 2, Annex 1) is called for. We request the further use of the concept of transfer pricing mechanism. Firstly, the definition of liquidity costs proposed by CEBS is reasonable. It covers not only the normal costs of funding, but also the indirect costs which may arise from procuring additional liquidity in the event of a liquidity bottleneck. Indirect costs may be incurred, inter alia, from holding liquidity buffers. In addition, CEBS distinguishes between marginal and average marginal costs of funding. Against the background of this very wide definition, we point out that a stringent application of the liquidity allocation mechanism cannot extend equally to all direct and indirect costs. In this respect, we advocate a sharper differentiation within the Guidelines between the individual liquidity cost categories. For example, it cannot be ascertained in every case that in addition to the direct costs of funding, the indirect liquidity costs are also considered in the product calculation. The definition of the marginal and average marginal costs of funding is not comprehensible. In particular, in connection with the requirement in point 25, it does not give sufficient explanation for the calculation of transfer prices. We request clarification. Otherwise, we suggest renouncing both concepts. We expressly welcome the reference to the proportionality principle. However, in the application of this principle, it should not only be the business model or the size of the institution which play a role. It is precisely in connection with the identification of liquidity costs that the funding situation in particular should also be decisive. Accordingly, the wording of sentence 2 of point 3 (page 3) could read: Respecting the proportionality principle, they are intended to apply to a wide range of institutions in terms of size, business model and funding structure (also see point 18). Moreover, there is a lack of concrete pronouncements in the consultation paper on how the proportionality principle can be applied on the basis of risk in a transfer pricing mechanism for the liquidity risk. We support the flexibility of implementation of the Guidelines proposed by CEBS. It gives the institutions the necessary scope to consider institution-specific particularities in the implementation. Furthermore, we should like to point out that full implementation of the concept of liquidity costs may have far-reaching effects on the lending of the institutions.
- 3 - Also in reaction to the financial market crisis, the vast majority of credit institutions have already established procedures to allocate the costs of liquidity in their internal processes and to consider them in the pricing process. In spite of that, the implementation of the Guidelines will require considerable adjustments in many institutions. The planned application of the Guidelines from 30 March 2011 therefore seems to us to be very ambitious. Instead, we propose application of the Guidelines only from 1 January 2012. The following comments relate mainly to the five Guidelines of the consultation paper. Specific observations Guideline 1 - Liquidity management framework Point 2 We request the deletion of the reference made in point 2, according to which especially large institutions have a transfer pricing system that is used for product calculation or to assess net interest income. It could give rise to the impression that only big banks deploy or have to deploy such a mechanism. Point 4 According to point 4, an institution must define its risk tolerance. We request clarification of the fact that this requirement in this connection relates only to liquidity risks. The wording should accordingly read: Institutions must have a clear definition of liquidity risk tolerance. In this way it is ensured that the concept of risk tolerance does not relate directly to the liquidity costs. Guideline 2 - Governance structure Point 7 The liquidity cost benefit allocation mechanism is to be controlled and monitored in order to legitimise and justify the internal prices derived with the help of this mechanism for the business areas using them. In our opinion, this provision is totally unclear. On the one hand, it is not obvious who is to undertake the control. On the other, it is not clear what the difference is between control and monitoring. Furthermore the purpose of the monitoring (or control) is unclear. Last but not least, it is not clear whether there is a difference between legitimation and justification of internal prices. We interpret this provision as meaning that the process of transfer pricing must be actively controlled in order to ensure proper costs or benefits for the liquidity-demanding or liquidity-
- 4 - generating business units. We should therefore like to ask CEBS to formulate the requirement more clearly and unambiguously. In this connection, we suggest replacing the term end-user business area by the concept liquidity users (in contrast to liquidity providers ). Point 9 The allocation mechanism based on the established risk tolerance, alongside other management tools, is to provide a tool for effective planning of the balance-sheet structure. In this respect, we should like to point out that the management of the balance-sheet structure not only is based on liquidity control, but also includes further control parameters (including capital or risk assets allocation). What is probably meant is that the allocation mechanism is to be used to plan the liquidity structure of the balance sheet. This should be clarified accordingly. On account of the thematic link, it would moreover be better to join point 9 to point 4 in Guideline 1. Point 10 According to CEBS the business areas charged with implementing and monitoring the internal prices should not be profit-oriented. Personnel working in these areas should not be set profit targets. The associated assignment of the liquidity management of the institutions is in principle to be welcomed. However, it is unclear how extensively this provision is to be interpreted. If the implementing of the internal prices is to be equated with pricing, the requirement would not necessarily lead to the desired results. As a rule, transfer prices are set by treasury units. These units often also have profit-centre or cost-centre guidelines and operate close to the market. This closeness to the market is also absolutely essential to set market prices or prices which can be derived from the market. The transfer prices are an instrument to pass on risks in this case liquidity risks within the institution so that these can be (centrally) controlled. These risks are controlled on the basis of a limit system and monitored independently of the market. Furthermore, a restriction, as called for in the Guidelines, is not appropriate and restricts the institutions in their organisational freedom. Point 11 We expressly welcome the call for consistent framework conditions (transfer pricing, policies) within a group of institutions. However, the implicit assignment of the liquidity management function to a central treasury unit would not be appropriate for every group of institutions. In general, the functions of the units involved in the liquidity management of a group of institutions (business strategy decision-making, control or monitoring) are not presented selectively in the consultation paper. We therefore suggest generalising the
- 5 - description of the function. Furthermore, transparency is necessary with regard not only to the balance sheet items, but also the off-balance-sheet items. Guideline 3 Use of the output from the allocation mechanism Point 12 According to point 12, the allocation mechanism should be designed to ensure that the endusers understand the output and know how to use it for decisions that will ultimately impact the financial situation of the institution. The reference to the financial situation of an institution seems ambiguous, as it suggests that the design of the allocation mechanism and the decision-making process is only to relate to such decisions. We therefore ask for the deletion of this reference. Furthermore, the internal prices should even percolate down to decision-makers at transaction level. In our view, this requirement to ensure the correct incentive effect can be dispensed with and what is more can be implemented only with difficulty. Especially for high volume trading desks, the focus is not on the funding of individual transactions undertaken in the course of the day, but on the liquidity costs associated with the entire position. In our view, allocation of the internal prices at business area level is sufficient to ensure that the business manager forwards the relevant cost components at transaction level to the decision-makers. This would also ensure that the trading area has a comprehensive understanding of the effects of the composition of the assets on their funding position. We therefore ask for the following reformulation of the second sentence: The internal prices should percolate down to business line decision makers to ensure maximum impact. Point 14 In point 14, there is a call for each funding operation to have an associated price. For this purpose, it should first be clarified that it is possible to aggregate deals with a similar liquidity profile. Furthermore, we understand the requirement to mean that the price associated with an operation does not have to contain all components of the CEBS liquidity cost concept directly. Moreover, the internal prices should be aligned with market transaction prices. Here, we first ask for clarification of what CEBS means here by market transaction prices. In this connection, we should like to point out that direct market prices are not available for all products offered in the commercial credit business. In contrast to interest hedges, there is no market for liquidity options. For this reason, most prices are derived for customers from several products.
- 6 - Guideline 4 - Scope of application Point 15 Point 15 focuses on the correct assessment of the deposits of an institution. We reject this explicit mention of deposits and the blanket call for pricing in the risk of withdrawal of deposits. It should be left to the institution to price in the indirect liquidity costs appropriately in accordance with its business model or its funding situation. In this respect, we ask for a more general formulation for point 15. Point 16 We agree with CEBS that the market liquidity risks and not the current contractual maturities should be reflected in the internal prices. However, in our opinion, the expected holding period and the market liquidity risk cannot be reflected simultaneously in the prices. Each of these parameters would lead to different prices. Whether the inclusion of the expected holding period or the market liquidity risk in the internal price for an asset is correct should be established for each transaction. For example, the expected holding period could be priced in for banking book assets or for traded loans, whereas the pricing for actively traded securities could focus on the market liquidity risk. Nevertheless, for traded securities which are to be held for longer periods than their market liquidity would require, consideration of the expected holding period in the internal prices seems appropriate. In our opinion, both the expected holding period and the market liquidity risk should be components of the transfer pricing model, but should not necessarily both have to be included in parallel in determining individual prices. Sentence 2 should therefore be reformulated as follows: The funding price charged should reflect either the expected holding period or the market liquidity risk (change in marketability). According to the ideas of CEBS, when determining the internal prices for marketable assets, haircuts should be made which reflect abrupt adverse changes in the liquidity of the markets on which these assets are traded. To this end, in the view of CEBS, the results of stress testing could also be used. In our opinion, the results of stress testing are not an appropriate starting point for the assessment of liquidity costs. This would not lead to prices in line with the market in normal market phases. Furthermore, the prices determined according to these specifications would lead to a double charge for the institutions, since the volume of the liquidity buffer to be held already takes this stress scenario into account. The last sentence should therefore be deleted. Point 17 For uncommitted credit lines, the business areas are to incur a charge in the same way as for committed credit lines. In our opinion, however, uncommitted credit lines are associated with
- 7 - a lower risk than committed credit lines and should therefore give rise to lower charges. During the financial crisis, differing conduct could be observed on the part of the institutions: unsecured uncommitted money market lines were sometimes cancelled. However, the uncommitted credit lines to non-banks were mostly retained for reputation reasons. These experiences show that uncommitted credit lines are not necessarily associated with the implicit promise of maintenance. Therefore, we propose the following new formulation of the third sentence: For credit lines with an evident implicit support, the business units granting the facilities. Credit institutions are regularly to be observed in practice also in fact adjusting committed credit lines depending on the credit standing of the borrower. Furthermore, we should like to point out that the majority of institutions currently charge no internal transfer prices for uncommitted credit lines. If all types of institutions were not to meet this specification, distortions of competition will result. The costs for this type of product in our view are to be borne by the customer. This could lead to uncommitted credit lines becoming more expensive. Points 19 and 20 The requirements of point 20 essentially represent concretisation of point 19. However, in our view, the specific statements in point 20 are unnecessary. We therefore ask for point 20 to be deleted in full. Moreover only the relevant liquidity costs should be taken into account. The wording of point 19 could read: The transfer prices should reflect - according to type, size and business model - current market and stressed funding conditions as well as the actual institution-specific circumstances, and should reflect both relevant direct and indirect costs, including the cost of a liquidity buffer 5 (see the example in Annex 2). Guideline 5 - Robust methodologies Point 23 To determine the internal pricing, the institutions should first use a market-determined interest rate structure curve. The indication that many institutions make use of EURIBOR or LIBOR curves for the calculation of floating-rate transactions and the swap cure for fixed-rate transactions seems to us to be arbitrary. The use of interest rates for the purposes of calculation is in practice too varied for the general example to better explain Guideline 5. We therefore request deletion of the example. Point 24 Product approval and the internal pricing process are to be integrated. This specification is unclear, in our view.
- 8 - Point 25 The definition of marginal is unclear to us in the context of the Guidelines, since the average rate does not always represent the marginal costs of the liquidity. For positions which are to be held to maturity, the current costs can be determined via a liabilities position with matching maturity. The current costs of funding the liabilities position should be reflected in the internal pricing of this position. On the other hand, for positions which are not to be held to maturity, a pricing model should be developed which measures the market liquidity risk. In this case, the costs of funding correspond to an average value of the currently existing costs of funding over various maturities. Although the homogeneous and the current costs of financing are to be considered in the transfer pricing mechanism, it does not lead to the desired results in our view to include both simultaneously in the determination of individual prices, since they represent different liquidity risks. We propose the following new formulation of sentences 1 to 4: The internal prices used should reflect the external cost of funding. The transfer pricing model should capture the cost of existing funding as well as the current funding costs. To achieve a reliable internal funding price, an institution's transfer pricing model must have the flexibility to adjust pricing variables to capture changes in costs for new funding, mainly, when calculating the contingent liquidity cost price. The call for a recalculation of the transfer prices after each new assets-side or liabilities-side transaction proves to be neither meaningful nor in line with practice. To control both new credit business and funding (deposits, securities issues), the institutions need a reliable basis of calculation. Customers expect a certain consistency of pricing over time. Control impetuses resulting from changed transfer prices therefore as a rule occur at longer intervals, with changes from one day to the next constituting the exception. From the technical point of view, the call for real-time adjustment of the transfer price would give rise to considerable expense. Accordingly, the last sentence should be deleted.
- 9 - Annex 1 - Liquidity cost allocation - examples Annex 1 describes the status of the implementation of transfer pricing in selected banks. However, no pronouncement is made on the extent to which the procedures discovered are in accordance with the new requirements or even possibly meet these in full. A benchmark implementation cannot therefore be inferred from this. From the scope of the paper, it can be concluded, however, that the new requirements are currently met in full by only a very few banks. It is therefore to be expected that in future banks will have to produce as a rule considerable one-off and continuous efforts in IT and human resources in order to implement the requirements. Possibly transfer pricing justifies a new function of its own in the organisation of the bank. CEBS is asked to pronounce on the general expectation of the Committee concerning the implementation efforts in banks. These should be graduated roughly accordingly to the degree of complexity of banks. Such expectations could make it easier for the banks to estimate the scale of the necessary work and to plan their own implementation. Besides, it is suggested dispensing with the entire Annex. We would appreciate it if our comments were taken into consideration in CEBS further deliberations. Yours faithfully ZENTRALER KREDITAUSSCHUSS p.p. Hartmut Kaempfer